refactor: Removes obsolete TRT-LLM request-abort disable wiring and re-enables...

refactor: Removes obsolete TRT-LLM request-abort disable wiring and re-enables cancellation test coverage (#8562)
Signed-off-by: Jacky <18255193+kthui@users.noreply.github.com>

components/src/dynamo/trtllm/backend_args.py

@@ -153,13 +153,6 @@ class DynamoTrtllmArgGroup(ArgGroup):
             default=False,
             help="If set, publish events and metrics to Dynamo components.",
         )
-        add_negatable_bool_argument(
-            g,
-            flag_name="--disable-request-abort",
-            env_var="DYN_TRTLLM_DISABLE_REQUEST_ABORT",
-            default=True,
-            help="Disable calling abort() on the TRT-LLM engine when a request is cancelled.",
-        )
         add_argument(
             g,
             flag_name="--load-format",
@@ -503,7 +496,6 @@ class DynamoTrtllmConfig(ConfigBase):
     extra_engine_args: str
     override_engine_args: str
     publish_events_and_metrics: bool
-    disable_request_abort: bool
     load_format: str
     model_loader_extra_config: str
     guided_decoding_backend: Optional[str] = None

components/src/dynamo/trtllm/request_handlers/handler_base.py

@@ -216,7 +216,6 @@ class RequestHandlerConfig:
     shutdown_event: Optional[asyncio.Event] = None
     generate_endpoint: Optional[Any] = None
     encoder_cache_capacity_gb: float = 0  # Encoder cache capacity in GB
-    disable_request_abort: bool = True
     additional_metrics: Optional["AdditionalMetricsCollector"] = None
     max_seq_len: Optional[int] = None
     disagg_machine_id: int = 0  # 10-bit machine_id for snowflake disagg_request_id
@@ -249,7 +248,6 @@ class HandlerBase(BaseGenerativeHandler):
         self.kv_block_size: int = config.kv_block_size
         self.shutdown_event = config.shutdown_event
         self.generate_endpoint = config.generate_endpoint
-        self.disable_request_abort = config.disable_request_abort
         self.additional_metrics = config.additional_metrics
         self.max_seq_len = config.max_seq_len
         self.disagg_machine_id = config.disagg_machine_id
@@ -464,13 +462,6 @@ class HandlerBase(BaseGenerativeHandler):
                 return_when=asyncio.FIRST_COMPLETED,
             )
 
-            # Abort the generation unless disabled
-            if self.disable_request_abort:
-                logging.debug(
-                    f"Request ID {context.id()} cancelled but abort() skipped "
-                    "(DYN_TRTLLM_DISABLE_REQUEST_ABORT=true)"
-                )
-            else:
             generation_result.abort()
             logging.debug(f"Aborted Request ID: {context.id()}")
 

components/src/dynamo/trtllm/tests/test_trtllm_handler_base.py

@@ -338,47 +338,6 @@ class _ConcreteHandler(HandlerBase):
         raise NotImplementedError
 
 
-class TestHandleCancellationAbortToggle:
-    """Tests for the disable_request_abort toggle in _handle_cancellation."""
-
-    def _make_handler(self, disable_request_abort: bool) -> HandlerBase:
-        """Create a HandlerBase with mocked config."""
-        config = MagicMock()
-        config.disable_request_abort = disable_request_abort
-        config.shutdown_event = None
-        return _ConcreteHandler(config)
-
-    @pytest.mark.asyncio
-    async def test_abort_called_by_default(self):
-        handler = self._make_handler(disable_request_abort=False)
-        generation_result = MagicMock()
-        context = MagicMock()
-        # async_killed_or_stopped returns an awaitable that resolves immediately
-        # (simulating the client cancelling the request)
-        killed_future = asyncio.get_event_loop().create_future()
-        killed_future.set_result(None)
-        context.async_killed_or_stopped.return_value = killed_future
-        context.id.return_value = "test-id-1"
-
-        await handler._handle_cancellation(generation_result, context)
-
-        generation_result.abort.assert_called_once()
-
-    @pytest.mark.asyncio
-    async def test_abort_not_called_when_disabled(self):
-        handler = self._make_handler(disable_request_abort=True)
-        generation_result = MagicMock()
-        context = MagicMock()
-        killed_future = asyncio.get_event_loop().create_future()
-        killed_future.set_result(None)
-        context.async_killed_or_stopped.return_value = killed_future
-        context.id.return_value = "test-id-2"
-
-        await handler._handle_cancellation(generation_result, context)
-
-        generation_result.abort.assert_not_called()
-
-
 class TestDeferredAbortGuard:
     """Tests for _DeferredAbort in disaggregated decode cancellation.
 
@@ -388,9 +347,8 @@ class TestDeferredAbortGuard:
     that waits for the first token before calling real abort.
     """
 
-    def _make_handler(self, disable_request_abort: bool = False) -> HandlerBase:
+    def _make_handler(self) -> HandlerBase:
         config = MagicMock()
-        config.disable_request_abort = disable_request_abort
         config.shutdown_event = None
         return _ConcreteHandler(config)
 
@@ -456,7 +414,7 @@ class TestDeferredAbortGuard:
     @pytest.mark.asyncio
     async def test_no_guard_in_non_disagg_mode(self):
         """Without _DeferredAbort wrapper, abort fires immediately on cancel."""
-        handler = self._make_handler(disable_request_abort=False)
+        handler = self._make_handler()
         generation_result = MagicMock()
         context = MagicMock()
         killed_future = asyncio.get_event_loop().create_future()
@@ -472,7 +430,7 @@ class TestDeferredAbortGuard:
     @pytest.mark.timeout(5)
     async def test_shutdown_calls_abort_directly(self):
         """Shutdown calls abort on whatever is passed (wrapper or real), immediately."""
-        handler = self._make_handler(disable_request_abort=False)
+        handler = self._make_handler()
         handler.shutdown_event = asyncio.Event()
 
         # Pass a _DeferredAbort wrapper — shutdown should still call .abort()
@@ -497,20 +455,6 @@ class TestDeferredAbortGuard:
         # Shutdown calls guard.abort() → since no first token, spawns background task
         # The important thing is EngineShutdown is raised and abort path is entered
 
-    @pytest.mark.asyncio
-    async def test_disable_request_abort_skips_guard(self):
-        """When disable_request_abort=True, abort is never called (guard irrelevant)."""
-        handler = self._make_handler(disable_request_abort=True)
-        generation_result = MagicMock()
-        context = MagicMock()
-        killed_future = asyncio.get_event_loop().create_future()
-        killed_future.set_result(None)
-        context.async_killed_or_stopped.return_value = killed_future
-        context.id.return_value = "test-disabled"
-
-        await handler._handle_cancellation(generation_result, context)
-        generation_result.abort.assert_not_called()
-
 
 class TestMultimodalGuard:
     """Tests for multimodal guard when --modality multimodal is not configured."""

components/src/dynamo/trtllm/workers/llm_worker.py

@@ -593,7 +593,6 @@ async def init_llm_worker(
             kv_block_size=config.kv_block_size,
             shutdown_event=shutdown_event,
             encoder_cache_capacity_gb=config.multimodal_embedding_cache_capacity_gb,
-            disable_request_abort=config.disable_request_abort,
             additional_metrics=additional_metrics,
             max_seq_len=config.max_seq_len,
             disagg_machine_id=int(endpoint.connection_id()) % 1021,

tests/fault_tolerance/cancellation/test_trtllm.py

@@ -40,7 +40,6 @@ pytestmark = [
     pytest.mark.model(FAULT_TOLERANCE_MODEL_NAME),
     pytest.mark.nightly,
     pytest.mark.parametrize("request_plane", ["nats", "tcp"], indirect=True),
-    pytest.mark.skip(reason="Cancellation is temporarily disabled"),
 ]
 
 
@@ -364,6 +363,7 @@ def test_request_cancellation_trtllm_decode_cancel(
                 )
 
 
+@pytest.mark.skip(reason="TRT-LLM prefill cancellation is disabled due to reliability")
 @pytest.mark.timeout(195)  # 3x average
 def test_request_cancellation_trtllm_prefill_cancel(
     request, runtime_services_dynamic_ports, predownload_models